Predictive method of S surface control using on SVM for AUV

Zhang Guocheng,Wan Lei,Li Yueming
DOI: https://doi.org/10.13245/j.hust.2012.03.014
2012-01-01
Abstract:A predictive S surface controller based on support vector machine(SVM) was obtained to analyze typical nonlinearity control system of torpedo-shaped autonomous underwater vehicles(AUV).The control algorithm was based on the S surface controller,which imported the idea of prediction control and combined with the excellence of support vector machine.Firstly the model of AUV was established to predict the motion states by the SVM.Then the control parameters of S surface controller were optimized by the quadratic performance function.At last the best control law was obtained.Simulation on AUV shows that the controller has merits of easy design,quick response,and excellent robustness,which is feasible and effective.
What problem does this paper attempt to address?